Acute upper respiratory illnesses (URIs) are a very common, important, and expensive clinical problem. The acute URI is the most common illness experienced by children and adults and is frequently complicated by bacterial infections such as acute otitis media or acute sinusitis especially in children. Two dilemmas for the clinician are predicting at the outset which URIs will become complicated by sinusitis and after having identified a likely episode, knowing which episodes of sinusitis will resolve spontaneously and not require antibiotic therapy. Here, we propose to capitalize on a large collection of respiratory specimens previously obtained from children with URI, based on the premise that analysis of selected nasal cytokines at 3 days will accurately predict those children at high-risk for developing sinusitis and at 10 days will predict the likelihood of spontaneous resolution versus the necessity for antibiotic therapy thus offering a useful biomarker of infection. We recently studied a cohort of 278 otherwise healthy 4- to 7-year-old children followed prospectively for one year to document their URIs and determine the incidence of sinusitis. Nasal wash samples were obtained on day 3 (acute URI) and day 10 (when the diagnosis of sinusitis was made) for the identification of virus and bacteria by PCR. Surveillance samples were also obtained 4 times/year when children were asymptomatic. Preliminary data suggest that in about 1/3 of episodes of presumed bacterial sinusitis a 2nd virus is identified on day 10. A pilot investigation of the cytokine profile of these nasal samples at 3 days shows dramatic differences between the samples in which virus is present compared to those in which virus is absent. Categorization of the cytokine signature in the 10-day sample will provide insight as to the cause of the persistent or worsening symptoms; if consistent with viral infection, antimicrobials may be avoided. Using nasal wash samples which have already been collected, we will sort the samples obtained into 4 cohorts: (A) symptomatic and uncomplicated URI in which a single virus was identified on day 3; (B) symptomatic and complicated URI (sinusitis patients) in which a single virus was identified on day 3; (C) symptomatic and uncomplicated URI in which no virus was identified on day 3 and (D) children with acute sinusitis on day 10. Our goal is to characterize the cytokine/protein signature for cohorts A-D for the following biomarkers: IFN?, IFN?, IFN?, IL-1?, TRAIL, IL-6, IL-8, IL-10, IP-10, TNF?, MCP-3, IL-28? and CRP. We will compare the concentrations and distribution of cytokines and CRP in each group to determine if (1) there is a characteristic cytokine signature for virus infections; (2) we can predict the development of complications in group B and (3) the cytokine/protein signature on day 10 in Group D will permit discrimination between a 2nd viral infection and bacterial superinfection. There is great potential to identify children with clinically defined sinusitis for whom antibiotics can be safely withheld thereby avoiding cost, adverse effects, and the emergence of antibiotic resistance.